Gastric Cancer Biomarkers Identified through Machine Learning Analysis

Strategic Insights -

Researchers at Lanzhou University have made significant progress in identifying potential biomarkers for early gastric cancer diagnosis. Through a machine learning-based approach, the team identified several genes associated with immune cell infiltration in the tumor microenvironment. The new biomarkers, including TAGLN2, HSP90AB1, SH3BGRL3, and CFL1, demonstrated strong diagnostic performance and were validated through quantitative reverse transcription polymerase chain reaction and immunohistochemical staining. The findings have important implications for improving early diagnosis and individualized treatment strategies for gastric cancer.

Key Takeaways:

  • Researchers at Lanzhou University identified four potential biomarkers for early gastric cancer diagnosis, including TAGLN2, HSP90AB1, SH3BGRL3, and CFL1.
  • The biomarkers were associated with immune cell infiltration in the tumor microenvironment and demonstrated strong diagnostic performance.
  • Machine learning models, including the glmBoost + XGBoost model, achieved high accuracy in diagnosing early gastric cancer.
  • The study validated the expression of candidate genes in GC tissues and adjacent non-cancerous tissues using quantitative reverse transcription polymerase chain reaction and immunohistochemical staining.
  • The research was funded by the National Natural Science Foundation of China and the Gansu Province Health Industry Scientific Research Project.
  • The study's lead author, Kewei Du, is available for comment and further information.

Statistics:

  • Mean AUC value for the glmBoost + XGBoost model incorporating B2M, CFL1, CTSD, and HSP90AB1 was 0.792.
  • 101 algorithm combinations achieved an average AUC of 0.7.
  • The nomogram integrating gene expression and clinical data was validated through calibration and decision curve analyses.
  • The study analyzed serum samples from 107 GC patients and 107 healthy controls.
  • The researchers employed single-cell RNA sequencing and immune infiltration analysis to evaluate the relationship between gene expression and immune cell function.

Sources:

  • NewsRx. Data from Lanzhou University Provide New Insights into Gastric Cancer (Identification of multiomics and immune infiltration-associated biomarkers for early gastric cancer: a machine learning-based diagnostic model development study). Cancer Weekly. June 17, 2025; p 909.
  • "Identification of multiomics and immune infiltration-associated biomarkers for early gastric cancer: a machine learning-based diagnostic model development study." BMC Cancer, 2025,25(1):1-23.
  • BMC Cancer. http://bmccancer.biomedcentral.com
  • https://doi-org.sdpl.idm.oclc.org/10.1186/s12885-025-14396-2